Blood gas poisoning with carbon monoxide (CO) remains a major cause of death and disability, affecting 50,000 persons a year in the U.S. alone. Patients with greater than 30% carboxyhemoglobinemia may develop brain injury, long-term neurocognitive deficits, and/or death. In the present proposal we introduce the first antidotal therapy for carbon monoxide poisoning, based on the finding of a surprising and near-irreversible CO-binding affinity of mutationally engineered human neuroglobin. Neuroglobin, like cytoglobin and many plant Hbs, is a six-coordinate hemoprotein, with the heme iron coordinated by a proximal and distal histidine residues. In an effort to understand the function of this bis-histidyl structure, we performed mutagenesis of the proximal histidine 64 molecule (e.g. H64Q) which opens the heme pocket, forming a five-coordinate molecule, more similar to the heme pocket of Hb or myoglobin. Unexpectedly, this molecule exhibits a remarkably high affinity for gaseous ligands, with a P50 value for oxygen (PaO2 value at which 50% of the heme binds oxygen) of 0.01 mm Hg (normal value is 26 mm Hg). This finding informs our primary hypothesis, that H64 mutant neuroglobin will bind CO with high affinity, producing a five-coordinate trap for CO. In our preliminary experiments we find that mutationally engineered H64Q neurglobin binds CO with an affinity more than 300 times that of hemoglobin and can remove CO from intact red cells in vitro and in vivo in CO exposed mice in less than 2 minutes. We propose to refine this molecule in a series of in vitro, pre-clinical physiology, and translational studies as a human biological antidote for CO poisoning and test its ability to clear CO from red blood cells, critical organs (lung, brain, and heart), and cytochrome C oxidase of the mitochondrial electron transport chain. From a translational perspective, will test the ability of this antidote to improve long-term organ function and survival after severe CO poisoning.